A minimum classification error method for face recognition

In this paper, we propose a minimum classification error rate based face recognition system. In our work, the minimum classification error formulation is incorporated into a neural network classifier called a multilayer perceptron. Experimental results show that our system is robust to noisy images and complex background.